When electronic technology is getting more and more advanced, machine learning and machine recognition are still difficult problems to be solved. Computers can calculate much faster than human being or memorize much more data than human being, but it is still difficult for computers to do what normal people can do very well, e.g. to recognize objects in an image.
But, scientists and engineers are still working day and night, trying to discover how to build machines that can do better machine vision, machine learning and even machine thinking because such topics can lead to great advancement of new technologies that will dramatically change human life.
Among these topics, machine recognition for images that are taken in earth-like environment is a very important field. With a better machine recognition for such images, various new applications may be built thereupon.
Therefore, it is beneficial to improve machine recognition and establish effective technical solutions thereof.